256 BELL SYSTEM TECHNICAL JOURNAL 



tion, if this range is exceeded they are lost and so it is necessary that 

 it be used within suitable limits. 



These materials have been described in technical papers before 

 various scientific societies and are not, therefore, discussed in detail 

 here. 



As indicating the wide scope of magnetic performance that is de- 

 manded of materials for use in communication apparatus, some of the 

 necessary properties are listed below : 



High permeability — at very feeble and at high inductions. 



High saturation value of induction. 



Low residual magnetization. 



Low hysteresis loss at feeble and moderate magnetizations. 



Low eddy-current losses over the frequency range from to 80,000 

 cycles. 



High constancy of permeability over a wide range of magnetization. 



Small effect on A.C. permeability at feeble currents with superposed 

 or residual D.C. magnetization. 

 Certain of these requirements are imposed from the simultaneous 

 transmission of D.C. telegraph currents, speech currents and carrier 

 frequency telephone or telegraph currents through the transformers, 

 loading coil or other iron-core apparatus in the circuit. Interference 

 between channels, due to magnetic modulation in the cores, must be 

 kept at an extremely low value for satisfactory quality of transmission. 



Summing up our work on materials, the results have been along two 

 general lines: (1) improvement in quality of commercial materials and 

 (2) development of new materials. As regards the first, we have 

 worked in close cooperation with material suppliers whose progressive 

 attitude has made possible certain of the advances described. The 

 more striking advances have been due to the discovery of new or im- 

 proved materials in our laboratories, the savings from which have 

 amply justified the program of continuous research which has been the 

 Bell System policy for a number of years. To take a single instance, 

 the field of magnetic alloys — probably the first to which we applied 

 intensive effort, — a single invention, the powdered electrolytic iron 

 core resulted in savings of such magnitude as to far overshadow the 

 cost of the investigational work. As already noted, this material has 

 since been superseded by the powdered permalloy core which represents 

 an equally great advance. 



There is one point which should be emphasized and that is, that the 

 most economical material is not necessarily the cheapest one. Treated 

 textiles cost more per pound than ordinary textiles; permalloy costs 

 more per pound than silicon steel. In these particular instances so 



